(ECEN 4018/5018) Game Theory and Multiagent Systems

This course provides an overview of game theory with a special emphasis on its application to multiagent systems, i.e., systems that are comprised of a collection of interacting and possibly competing decision making entities. Examples drawn from engineered, economics, and social models, including multivehicle robotics, data networks, sensor networks, and electronic commerce.

(ECEN 5028) Dynamic Programming

This course will cover the fundamentals of dynamic programming which is a method for solving complex problems by breaking them down into simpler subproblems. Topics include deterministic and stochastic formulation of the principle of optimality, value and policy iteration, introduction to finite state Markov chains, partial state information problems, stochastic shortest path problems, infinite horizon problems, and introduction to approximate dynamic programming. Applications include inventory control, finance, routing, and sequential hypothesis testing.

(CS/SS 241) Introduction to SISL: Algorithmic game theory

Over the last few years there has been enormous activity at the interface of computer science, game theory, economics, and control. In this course, our goal is to survey some important of the important new areas that are emerging in this field. Some of the topics we will study include:

This course is intended for graduate students and will be organized as a topics course. Post-docs are also encouraged to attend lectures on topics of interest to them, and need not be registered to do so. Registered students will be expected to present multiple lectures in addition to completing homework assignments. It is expected that students are comfortable with the basics of game theory, graph theory, probability theory, and Markov chains. (Co-taught with Adam Wierman and John Ledyard).

(CS/SS 241) Queueing Network Games

Queueing theory has long served as a fundamental tool for understanding the dynamics of computer systems from computer networks, to production systems, to airline scheduling, and beyond. However, traditional results view customer behavior as an exogenous parameter, unaffected by the details of the model, and thus cannot capture the impact of pricing and competition within the models. On the other hand, in recent years game theoretic techniques have been applied to the same range of problems in order to characterize the impact of customer behavior/reactions. However, using game theoretic techniques alone ignores the queueing dynamics inherent in these applications, e.g. networking.So, there is a mutual need for studying the interactions of game theory and queuing models. Our goal in this reading group is to quickly ramp up on the known literature combining game theory and queueing and then to develop and study new models in this area. To do this, we will work through the best (and only) book on the topic along with a number of important papers in the area. (Co-taught with Adam Wierman).